DesignBuilder Case Study

Performance Optimization of a Low-Temperature Chilled Water

Plant Serving a Suite of Operating Rooms

DesignBuilder’s Energy Management System (EMS) functionality was used to optimize the performance

of a low-temperature chilled water plant serving a suite of operating rooms (ORs). As a result of this

study, AKF was able to identify a potential controls adjustment that would result in an annual cost savings

of up to $20,000 for the low-temperature OR HVAC scope.

By Michael Sweeney, BEMP, LEED AP

At a Glance

An energy model was developed using DesignBuilder EnergyPlus to optimize the annual cooling

performance of several low-temperature air handlers and the chilled water plants serving them. The air

handlers serve a suite of fifteen ORs, and each room requires a high volume of supply air at a low

temperature setpoint.

The new operating suites are located within a mixed new/existing healthcare facility on Penn Medicine

Chester County Hospital’s campus. In addition to the design and modeling of the entire facility, the unique

air and thermal requirements of the OR suites warranted a separate focused study.

DesignBuilder's EnergyPlus-based EMS system allows the energy modeller to introduce highly-specific

and novel controls strategies that would otherwise be unavailable within traditional hourly simulation

engines or static energy calculations. As a result of this functionality, AKF was able to identify a potential

controls adjustment that would result in an annual cost savings of up to $20,000 for the low-temperature

OR HVAC scope.

DesignBuilder Case Study

DesignBuilder model used in the analysis

Project Scope

• Mixed new/renovation healthcare facility within existing campus, 236,000 total SF

• Low-temperature OR scope is approximately 20,000 SF

• In addition to the low-temperature ORs, there is a varied healthcare program including a bed tower,

exam room, and lobby/waiting space

• Location: West Chester, PA

Design Concerns

• The OR usage varies throughout the time of day and week, and air handlers are permitted to vary

down when a room is not in use. This results in significant part-loading effects that lower the annual

performance of the air-cooled chiller.

• An initial design was developed with each air handler serving ORs consisting of two chilled water coils

in series: the first coil is fed by the campus chilled water loop and pre-cools the air to 48oF. The second

coil is fed by the air-cooled chiller and brings the air down to 43oF. The goal of this strategy was to

take as much conditioning from the campus loop as possible, and then satisfy the remainder with the

less-efficient air-cooled chiller.

• During the design phase, the energy model was used to validate the setpoint used for the campus

cooling coil to optimize the performance of the air-cooled chiller throughout the year.

Energy Modeling Parametric Analysis Goals

• Run several iterations with varying pre-cool setpoints in order to analyze the impact on the air-cooled

chiller’s performance.

• Develop a practically feasible BMS routine to optimize the air-cooled chiller’s part-load efficiency by

varying the campus coil’s discharge temperature.

DesignBuilder Case Study

Project Details

Low-temperature ORs

• Cooling Setpoint: 65oF

• Air Change Requirement: 30 air changes per hour (ACH)

• Supply Air Temperature (SAT) Setpoint: 43oF

• Above setpoints and ACH requirements relax in an OR when it is not in use

Low-temperature Air Handlers

• Three 30,000 cfm air handlers serve the rooms, each equipped with (two) separate chilled water coils.

• Campus chilled water coil pre-cools the air and a second chilled water coil, supplied by a low-

temperature air-cooled chiller, cools the air to the 43oF setpoint.

• Air handlers are permitted to vary down when individual ORs are not in use.

• A series of models were developed to determine the ideal pre-cool setpoint for the campus chilled

water coil, and DesignBuilder’s EMS interface was used to develop a load optimization strategy to

optimize the air-cooled chiller’s performance.

Chilled Water Plant

• Campus Chilled Water Plant (existing): 42oF water temp, serves entire campus loop.

• Air-cooled Chiller Plant (new): 110 tons, 36oF water temp designed for a coil supply air discharge of

43oF. Chiller equipped with variable speed drive (VSD) operation and a glycol loop. Serves only low-

temperature air handlers.

Energy Modeling Approach

General Approach and Analysis

EnergyPlus was initially chosen for the model over eQuest due to the ability to model dual cooling coils

within a single air handler. The model and all iterations were developed entirely within DesignBuilder

using their Detailed HVAC interface and supplemented by their integrated EMS module.

Chilled Water Plant Efficiencies

The following scatter plots show the part load ratio (PLR) vs. coefficient of performance (COP) for both

the campus cooling plant and the low-temperature air-cooled chiller plant. These plots are based on

actually modeled results. See the Appendix for further detail on how the central plant loads and

efficiencies were modeled.

DesignBuilder Case Study

As can be seen in these two charts, both chiller plants operate at their peak efficiency in the 30%-45%

range. However, due to the sheer size of the campus plant in comparison to the OR scope, it is unlikely

that the OR sequence will significantly impact the campus plant’s PLR at any given point in time. In

contrast, the new air-cooled chiller plant’s PLR is entirely dependent upon the portion of low-temperature

air handler load satisfied by the upstream campus cooling coil. For this reason, the campus cooling coil’s

setpoint was chosen as the primary optimization variable.

The following design iterations were developed:

1. Air-cooled chiller only, satisfying entire load (two chillers identical to proposed design assumed to

satisfy entire load)

2. Campus cooling to 54oF (SAT) with air-cooled chiller satisfying remainder of load

3. Proposed design: campus cooling to 48oF (SAT) with air-cooled chiller satisfying remainder of load

4. Optimized Load Management: modulate campus cooling setpoint in order to optimize air-cooled

chiller part loading

-

1

2

3

4

5

6

0% 20% 40% 60% 80% 100%

Co

eff

icie

nt o

f P

erf

orm

an

ce

(CO

P)

Part Load Ratio (PLR)

Air-cooled Chiller COP vs Part Load

-

1

2

3

4

5

6

7

8

9

- 0.20 0.40 0.60 0.80 1.00 1.20Co

eff

icie

nt o

f P

erf

orm

an

ce

(CO

P)

Part Load Ratio (PLR)

Campus Chiller Plant COP vs Part Load

DesignBuilder Case Study

Energy Model Parametric Results

As expected, the initial design’s low campus precool option compared favorably to the air-cooled chiller

only and high campus precool iterations. However, the Optimized Campus Precool strategy led to savings

of roughly $2,000 per year or ~2.5% of total OR HVAC cost. This is not an enormous savings and it remains

to be seen if the practical and financial hurdles associated with controls programming are worth the cost

savings.

HVAC Cost

Summary ACC Only

54oF Campus

Precool

48oF Low Campus

Precool

Optimized

Campus Precool

Cooling-Elec $94,914 $71,409 $42,228 $ 39,418

Cooling - Campus $0 $18,280 $32,785 $ 32,548

Pumps - Total $1,426 $2,031 $1,672 $ 2,056

Total $96,340 $91,720 $76,684 $ 74,022

DesignBuilder Case Study

The graph above shows the hours per year at each part load for the low-temperature chiller. The green

band is the “sweet spot” of best efficiency. As can be seen, the optimized setpoint routine shifts

the most operational hours into this area.

Parametric Heat Maps

The following heat maps show the number of hours at each part-load and COP range for each design

iteration. Darker green indicates more hours at a given point of operation; lighter green indicates fewer

hours. A higher COP indicates better chiller performance.

Heat Map of COP vs. PLR: Air-cooled Chiller Picking up Entirety of OR Load

~2,400 hrs/year are spent at a COP above 4.5, resulting in an average annual COP of 2.5 for the air-cooled

chiller.

Heat Map of COP vs. PLR: Campus Chiller Plant Only Precools to 54oF

~3,100 hrs/year are spent at a COP above 4.5, resulting in an average annual COP of 3.4 for the air-cooled

chiller.

DesignBuilder Case Study

Heat Map of COP vs. PLR: Campus Chiller Plant Precools to 48oF

~3,000 hrs/year are spent at a COP above 4.5, resulting in an average annual COP of 3.0 for the air-cooled

chiller.

Heat map of COP vs. PLR: Optimized Load Management

~5,400 hrs/year are spent at a COP above 4.5, resulting in an average annual COP of 4.1 for the air-cooled

chiller.

DesignBuilder Case Study

Summary

DesignBuilder’s simulation results and its built-in EMS functionality provided the tools required to make

fully-informed plant optimization decisions that had real-life design impacts.

The ability to quickly take a previously-developed LEED model, strip out the important components and

run HVAC iterations quickly in DesignBuilder’s detailed HVAC mode afforded the design team a

confidence in the performance and controls of a complex cooling strategy that otherwise would not be

available with a static model based on load calculations.

DesignBuilder helped the design team validate their existing calculations and also provided ways to yield

more real-life operational cost savings throughout the year.

Michael Sweeney, BEMP, LEED AP is a mechanical engineer with In Posse, AKF’s in-house high performance building lab. Michael

excels in energy modeling and documentation for LEED® and local incentive programs in New York and New Jersey, such as NYSERDA Flex Tech and NJ Clean Energy P4P. He has experience with CFD modeling using Six Sigma, as well as working knowledge

of Python. Michael is a strong team player and is admired for his positive attitude, his excellent organizational skills, and his

diligence in accomplishing project goals.

| msweeney@in-posse.com | T: 212.626.0105 | akfgroup.com | in-posse.com |

In Posse provides consulting and engineering design services for high

performance, deep green projects with a special expertise in net zero

energy. At In Posse, we focus exclusively on the energy and sustainable

design sectors of the built environment for clients in a broad range of market

sectors. Our services address all aspects of building performance including modeling and analysis of

building systems, engineering high performance building systems, commissioning, and occupant

education and engagement. In Posse is a subsidiary of AKF Group and is headquartered in Philadelphia,

PA with an office in New York, NY.

DesignBuilder Case Study

Appendix – Inputs and Details

A. DesignBuilder image of one of three 30,000 cfm air handlers serving the suite of ORs: the first

cooling coil at each air handler is supplied by the campus chilled water plant and precools the air

prior to being run through the low-temperature cooling coil supplied by the air-cooled chiller

B. Screenshot of the EMS code used to vary the campus cooling coil supply air setpoint based on the

air-cooled chiller’s part load

DesignBuilder Case Study

C. OR Load Assumptions:

• Equipment Power Density: 4 watts/SF

• Lighting Power Density: 0.7 watts/SF

• Occupancy: 200 SF/person

D. The below table shows the variation assumed in the OR suites throughout the day/week and the

corresponding total supply air requirements supplied by each of the three air handlers.

Fraction of Full Load

Daytime

(8am – 8pm)

Evenings

(8pm-Midnight)

Overnight

(Midnight-8am)

Weekends

(All Hours)

AHU-01 100% 100% 100% 60%

AHU-02 100% 60% OFF OFF

AHU-03 100% OFF OFF OFF

Total CFM Delivered to Suite 90,000 48,000 30,000 18,000

E. Modeling the Campus Chiller Plant:

The campus chilled water plant was simulated in a separate energy model using an hourly load profile

developed using actual metered load data. The results of this simulation were used to calculate the

approximate cost per BTU of campus cooling load on an hourly basis. Because at any given hour the

portion of load satisfied by the air-cooled chiller vs. the campus plant may change, it was important to

consider if there are any hours when the optimization strategy used was not ideal in comparison to the

hourly performance of the campus plant.

DesignBuilder Case Study

Annual Load Profile of Air-cooled Chiller at 48oF Campus Precool Conditions:

-

200

400

600

800

1,000

1,200

1,400

1,600

1,800

1

382

763

114

4

152

5

190

6

228

7

266

8

304

9

343

0

381

1

419

2

457

3

495

4

533

5

571

6

609

7

647

8

685

9

724

0

762

1

800

2

838

3

Ce

ntr

al P

lan

t Lo

ad (T

on

s)

Hour of Year

Central Cooling Plant Annual Load Profile

0

10

20

30

40

50

60

70

80

90

100

13

667

311

096

146

11

826

219

12

556

292

13

286

365

14

016

438

14

746

511

15

476

584

16

206

657

16

936

730

17

666

803

18

396

Air

-co

ole

d C

hill

er L

oad

(To

ns)

Hour of Year

Air-cooled Chiller Annual Load Profile